Some embodiments described herein relate generally to aggregation of data units at access points, and, in particular, to methods and apparatus for self-tuning aggregation of data units at access points.
Some known aggregation approaches aggregate as many queued data units as possible during a time period for data transmissions between an access point and associated stations. Such known aggregation approaches, however, typically do not wait for the opportunity to aggregate more data units when the queued data units are few within the time period. As a result, the performance and channel utilization for these known aggregation approaches can be poor due to collision avoidance characteristics and transmission overhead associated with the aggregation.
Some other known aggregation approaches aggregate data units only when a minimum number of data units are queued for a data transmission between an access point and an associated station. Such known aggregation approaches can also implement a time-out mechanism to avoid long latency. Choosing a reasonable and effective criterion for the minimum number of queued data units and the time-out threshold, however, is typically difficult. Moreover, the chosen criterion is typically fixed and cannot be suitable for every station under every operating condition. As a result, the performance and service quality for some associated stations and services under some operating conditions may be degraded.
Accordingly, a need exists for a data unit aggregation approach that enables dynamic tuning of the aggregation setting for each associated station and under different operating conditions to provide better service, performance, and/or channel utilization for data transmissions.